A Study on the Optimization of Classroom Observation and Evaluation in Primary Mathematics Based on the Improved Flanders Interaction Analysis System

Abstract

Aiming at subjective judgment and ambiguous evaluation criteria prevailing in current lesson observation and evaluation of primary mathematics, this study takes the Improved Flanders Interaction Analysis System (IFIAS) as the core quantitative tool. Adopting a mixed research method integrating classroom observation, questionnaires and in-depth interviews, it constructs a segmented behavioral coding framework adapted to primary mathematics classrooms. A total of 118 mathematics lessons were observed and coded at 30-second intervals, generating 12,400 behavioral units. Questionnaire surveys collected 152 valid responses from participating teachers, and semi-structured interviews were conducted with 28 teachers. This paper systematically sorts out the necessity and feasibility of implementing this tool as well as its practical effects, and analyzes various obstacles existing in its application. It further constructs a closed-loop optimization path covering pre-class, in-class and post-class links, and puts forward supporting guarantee strategies from the perspectives of hierarchical training, digital empowerment and diversified teaching research, so as to provide operable practical references for primary and secondary schools to carry out data-oriented and professional mathematics teaching research.